Vietnamese Swimming and the Missing Data Layer Behind Every Race

I was sitting in the seventh row of a national swimming final, notebook open, pen resting on a blank page. When lane four touched the wall, the scoreboard flashed a single number. The stands applauded, the sound dissolving quickly in the humid air of the arena. I waited another fifteen seconds, waiting for the split line, for the five 50-metre marks, for stroke rate, for turn times. The board held that one number, bright and inert, like a door closing in front of anyone who wanted to step inside the race.

I turned to the coach beside me, a man who had stood on pool decks for more than twenty years. His answer was short: "Final time only." Those words followed me home, and they became the splinter I had to write out. In a final with ten lanes, we keep ten numbers. We drop hundreds of other pieces of information, and nobody on the officiating panel is assigned to pick them up.

My career started at the pool itself. In 2026 I joined Thanh Nien newspaper as a swimming reporter, sitting through sessions and recording every length by eye and by ear, because back then I had nothing else to lean on but direct observation. Seven years later I switched lanes: I moved into the technical analysis room of a football club, working with GPS data, expected-goals metrics and long spreadsheets. There I learned that one bad number can drag an entire conclusion down with it. A small GPS deviation was enough to teach me: verification is everything.

Since then I have carried one habit: no conclusion before the data has been cross-checked at least twice. When I returned to writing about swimming for the Vietnamese market, I kept that habit. The only difference is that here the problem is not that the data lies. The data is silent.

To see it clearly, place two pictures side by side. In every V.League football match, each player carries a tracking device on his back, and after the whistle the analysis room can read every metre run, every acceleration, every minute of recovery. At the pool, we have a scoreboard and a number. That contrast says something about how we invest in the two sports, and about where a sporting system can lose its trail.

I believe in numbers, but only after a number has passed three rounds of checking. With Vietnamese swimming, I have not cleared the first round.

At any international meet with proper standards, the organisers publish 50-metre splits for every finalist. That is the minimum data layer, not an added bonus. With it, an analyst can see how a race was paced: whether a swimmer went out fast or held back, whether there was a fade in the closing length, whether the third turn was a dead point. Without it, a race is reduced to two ends: the start and the touch. The entire middle — where tactics live — disappears.

The 50-metre split is the cheapest and most information-rich unit of data in swimming, and it is precisely what domestic meets leave behind.

Vietnam has had swimmers whose results should have built an entire data archive. Nguyen Thi Anh Vien closed her SEA Games career with 25 gold medals, a record for Vietnamese sport, and she competed at the London 2026 and Rio 2026 Olympics. Nguyen Huy Hoang won silver in the 800m freestyle at the 2026 Asian Games and appeared at the Tokyo 2026 Olympics in both the 800m and 1500m freestyle. Tran Hung Nguyen, Pham Thanh Bao, Hoang Quy Phuoc — each is a separate chapter, a separate body type, a separate way of distributing effort.

If their splits had been recorded and stored properly every time they entered the water, we would have the improvement curve of an entire generation. We have medals, but we do not have the curve. A medal is a photograph; a curve is a film. A sport built on photographs will always be guessing, and guessing does not accumulate.

Take one concrete example to show how large the loss is. In a 25-metre pool, a 200-metre race contains seven turns. Each turn, depending on technique, costs somewhere between six-tenths and nine-tenths of a second. Multiplied out, seven turns occupy a block of time large enough to decide the placing in any evenly matched race. For a sprinter like Hoang Quy Phuoc, the start and the turns are almost the entire race. Yet in domestic data, both of those things usually do not exist.

In the 400m individual medley the story is even clearer. The race splits into eight segments, each a different stroke. With only a final time, a coach cannot know whether the weakness sits in the breaststroke leg or the freestyle leg, cannot know at which segment the swimmer loses energy, cannot know whether to fix technique or fix effort distribution. One final number answers one question, while eight splits answer eight questions.

One detail stands out above the rest. In international data systems, when Vietnamese swimmers compete at continental or world level, their splits are recorded and stored in the organiser's database. That means data about them exists — but it sits elsewhere, held by someone else, and only a few times a year. Back home, where they swim dozens of times each season, where coaches watch them daily, the data vanishes the moment the board goes dark. We keep data at the meets least relevant to development, and we lose it at the meets most relevant.

Someone might ask why it is missing. The answer lies in operations, not attitude. Automatic timing systems are capable of exporting splits, but many domestic pools run systems that return only a final time, or have the equipment while nobody has been assigned to export it. A swim meet needs a timekeeper. A swim meet that wants data needs one more person: someone responsible for recording, checking and archiving. That role has not become a job title at most organising committees, and what has not become a job title usually has not become work.

The pandemic taught me to measure a league by its recovery index, not by its points. In 2026, when domestic football was suspended for months, I built a model from the GPS data of 365 players, combining high-intensity running distance, acceleration counts and injury history to estimate risk. When the league restarted, I predicted that the three teams pressing at the highest intensity would see injury risk rise by roughly twenty-three percent. My club cut training load by fifteen percent and lost no key players, while several other teams lost an average of three players each.

I tell that story not to show off a model, but to point out that the same logic applies to swimming. Swimming has its own injury profile: swimmer's shoulder, breaststroker's knee. Those injuries do not appear overnight; they accumulate from volume and intensity. To see them accumulate, you need a training log: volume, intensity zones, dryland work, the number of double sessions. No log, no model. No model, and every prevention measure becomes a reaction after the injury has already happened.

Based on my experience following domestic swim meets, the biggest gap sits in the habit of retention, not in the measuring equipment. We measure to learn who won. We have not yet measured to learn why.

Suppose tomorrow the splits were published in full. The largest change would come in selection. Today, picking a young athlete usually rests on the feel of an experienced eye, and that feel has value, but it cannot answer the question: a fourteen-year-old swimming a 200m individual medley with these splits — where does she sit in the distribution for her age group? Answering that requires a distribution. A distribution requires years of data. A selection model does not replace the coach's eye; it only narrows the number of people that eye has to examine.

A second change would come in how we read a record. Breaking a national record by going out far too fast and dying at the end is a very different signal from breaking it by accelerating over the final 150 metres. Both write an identical line in the record book. If I had to choose one metric to follow Vietnamese swimming over the next two years, I would choose the gap between the first half and the second half of each race.

In a championship cycle that pressure grows. As a SEA Games approaches, all attention converges on the medal count, and everything else is set aside. But medals at a Games are decided by what accumulated over the previous three years, mostly at domestic meets nobody broadcasts. A coach preparing for that cycle needs to know at which segment, at which distance, under which pool conditions his swimmer fades. Without data, he can only rely on memory, and memory is selective.

There is a temptation I have to warn myself against. When data becomes a slogan, people begin to believe that data inevitably brings medals. That correlation does not exist as a simple cause. Many strong swimming nations keep dense data, but they are strong because of their development systems, their money, their geography, their culture of retaining people. Data is an accompanying trace, not the engine. Mistaking the trace for the engine is a mistake I have seen many times.

I have also seen a correct number used badly. In 2026, a domestic club asked me to advise on a transfer. I analysed nineteen matches of the striker they wanted: he had scored eighteen goals but his expected-goals figure was 11.2, a conversion rate of 31.4 percent, nearly double the league average of 15 to 18 percent. Seventy percent of his goals came from set pieces, entirely dependent on the system. I recommended against the signing. The leadership ignored it, saying numbers could not replace the eye for a player. He scored four goals in twenty matches and suffered two hamstring injuries.

Both directions of that story are true: a number can save a transfer, and a number can be thrown into an argument it was never built to win. What I take from it is not blind faith in models, but a principle: every model must travel with its own limitations, its sample size, its assumptions and its error margins. I have never issued an absolute prediction, and I do not intend to start now.

And when the source material contains nothing, the most honest response is to say so. I received an analysis with every data field empty: no splits, no pool conditions, no athlete profile, not a single line of technical description. A writer could fill that void with emotional language, with miracles, with destiny. I refuse. Croatia 2026 was not a miracle — it was xG written into history. And a blank table must be read as a blank table.

Data does not tell stories; it records everything so that I can tell them myself. When there is nothing recorded yet, I have nothing to tell yet.

The signals worth watching in the next cycle are specific. Whether the most recent national swimming championship published 50-metre splits, and in what form — that is the easiest signal to verify. Then, whether any organising committee adds a data role, even part-time. And further out, whether the training plans of the major centres begin to include volume and intensity logs, alongside a health tracker for shoulders and knees.

A strong swimming nation does not begin with medals. It begins by accepting that after every race, something must be kept — a number, a split, a line of notes sufficient to spare whoever comes next from starting again at zero.